A transresistance amplifier is an active circuit that converts an input current into a proportional output voltage, functioning as a current-to-voltage converter with a gain expressed in ohms. What this changes in a real circuit is the ability to interface high-impedance current sources—like photodiodes, ionization chambers, or piezoelectric sensors—with low-impedance voltage inputs, such as microcontroller ADCs or oscilloscopes, without loading down the source and destroying the signal. People most commonly confuse it with a simple passive shunt resistor, or they get tripped up by the interchangeable (but technically distinct) terms 'transresistance' and 'transimpedance'.
Think of it like a toll booth (the amplifier) that counts individual cars (electrons/current) passing through and outputs a digital receipt with the total dollar amount (voltage). A simple shunt resistor is like a physical scale the cars drive over; if the scale is too heavy (high resistance), it causes a traffic jam (voltage buildup that alters the source's behavior). The active amplifier keeps the road perfectly flat (virtual ground) while still counting the cars.
The Core Math and a Worked Numeric Example
The fundamental transfer function for an ideal inverting transresistance amplifier is straightforward:
Vout = -Iin × Rf
Where Iin is the input current and Rf is the feedback resistor. The negative sign indicates phase inversion, which is easily corrected in software or by referencing the ADC to a mid-supply bias.
Let’s look at a real-world bench scenario. You are building a light meter using a Vishay BPW34 PIN photodiode. Under your target illuminance, the diode generates 5 µA of photocurrent. You need a 1.0V output swing for your 12-bit ADC, so you select a 200 kΩ feedback resistor.
- Calculated Output: Vout = -(5 µA) × (200,000 Ω) = -1.0 V.
On paper, this works perfectly. On the bench, you will likely see a heavily filtered, sluggish signal. Why? Parasitic capacitance. The BPW34 has a junction capacitance (Cj) of about 72 pF at 0V reverse bias. This capacitance sits in parallel with the input node, forming an unintentional low-pass filter with your 200 kΩ feedback resistor.
The -3 dB bandwidth limit is calculated as:
f-3dB = 1 / (2π × Rf × Cj)
f-3dB = 1 / (2π × 200,000 × 72 × 10-12) ≈ 11 kHz
Transresistance Amplifier vs. Simple Shunt Resistor
A common question from hobbyists is: 'Why do I need an op-amp? Can't I just put a 200 kΩ resistor in series with the photodiode and measure the voltage across it?' You can, but the performance degrades severely as gain increases. Here is how the two approaches compare in practice.
| Feature | Passive Shunt Resistor | Active Transresistance Amplifier |
|---|---|---|
| Input Impedance | High (Equal to Rshunt) | Extremely Low (Virtual Ground, ~0 Ω) |
| Voltage Burden | High (Alters diode bias point, causing non-linearity) | Near Zero (Diode remains at 0V or fixed bias) |
| Bandwidth Limit | Severely restricted by R × Cj time constant | Extended via Miller effect and active compensation |
| Noise at High Gain | Thermal noise of the resistor dominates | Resistor noise + op-amp voltage/current noise density |
The active topology maintains a constant voltage across the photodiode (usually 0V or a fixed reverse bias). This keeps the diode's junction capacitance constant and eliminates the non-linearities caused by the diode's own I-V curve bending as voltage builds up across a passive shunt.
Where You Meet Transresistance Amplifiers in Practice
You will rarely see a discrete transresistance amplifier built on a breadboard for high-speed applications; they are usually integrated or carefully laid out on custom PCBs. According to Texas Instruments' application data for the OPA657, a favorite FET-input op-amp for this topology, you will find these circuits in:
- LiDAR and Time-of-Flight Sensors: Avalanche photodiodes (APDs) generate tiny, nanosecond-scale current pulses. A high-speed TIA (using parts like the ADA4817 or OPA657) converts these to voltage for high-speed comparators. Safety note: APDs often require 50V to 100V reverse bias; treat these nodes as shock hazards.
- Pulse Oximeters and Heart Rate Monitors: Wearables like the MAX30102 use internal, programmable-gain transresistance amplifiers to read the microamp-level reflections of red and IR light off blood flow.
- Spectrophotometry and Chromatography: Laboratory equipment uses precision, low-noise TIAs to measure the absorption of light through chemical samples, requiring ultra-low input bias currents (femtoamp range) to avoid swamping the sensor signal.
Common Design Pitfalls and Edge Cases
If you are designing a TIA from scratch, avoid these three common mistakes that ruin circuit stability and noise performance.
- Ignoring the Feedback Capacitor (Cf): The input capacitance (diode Cj + op-amp Cin + stray PCB capacitance) creates a pole in the feedback loop, destroying phase margin and causing high-frequency oscillation. You must add a small feedback capacitor (often 1 pF to 5 pF) in parallel with Rf to introduce a zero that compensates the phase. Finding the exact value requires iterative bench testing or SPICE simulation.
- Misunderstanding Noise Gain: The signal gain of a TIA is Rf. However, the noise gain (which dictates op-amp stability and voltage noise amplification) is 1 + (Cin / Cf). If your input capacitance is 50 pF and your compensation capacitor is 1 pF, your noise gain at high frequencies is 51 V/V, even if your DC signal gain is only 10 V/V. This is why selecting an op-amp with high GBWP is non-negotiable.
- Using the Wrong Op-Amp Architecture: Bipolar input op-amps (like the classic LM358 or NE5532) have input bias currents in the nanoamp to microamp range. If your photodiode only generates 2 µA, the op-amp's bias current will introduce massive DC offset errors. Always use CMOS or JFET input op-amps for precision DC or low-frequency TIA applications.
Frequently Asked Questions
What is the difference between a transresistance amplifier and a transimpedance amplifier?
In pure academic theory, 'transresistance' refers to the real, DC component of the gain (measured in ohms, R), while 'transimpedance' refers to the complex AC gain (measured in ohms, Z, encompassing both resistance and reactance). In practical industry engineering, the terms are used interchangeably. The acronym TIA universally stands for Transimpedance Amplifier, even when the circuit is primarily passing DC current from a slow-moving sensor. If you are reading a datasheet or an All About Circuits tutorial, assume TIA covers both concepts.
How do I choose the right feedback resistor value for a photodiode TIA?
Start with your maximum expected input current and your target maximum output voltage (usually constrained by your ADC reference or op-amp supply rails). Use Ohm's law: Rf = Vout(max) / Iin(max). For example, if your ADC maxes at 3.3V and your brightest light condition yields 10 µA, Rf = 330 kΩ. Once you have the resistor value, verify that the resulting bandwidth (factoring in diode capacitance) meets your signal speed requirements. If the bandwidth is too low, you must lower Rf and add a secondary voltage gain stage later in the signal chain.
Why does my transresistance amplifier circuit oscillate at high frequencies?
Oscillation in a TIA is almost always caused by insufficient phase margin due to the input capacitance (photodiode junction capacitance plus op-amp common-mode input capacitance) interacting with the large feedback resistor. This creates a pole in the feedback factor (β) that causes the noise gain to spike at high frequencies. To fix this, add a small compensation capacitor (Cf) in parallel with the feedback resistor. A good starting point for Cf is calculated using the formula: Cf = √(Cin / (2π × Rf × GBWP)). You will likely need to tune this value on the bench using a network analyzer or by observing step-response ringing on an oscilloscope.
Can I use a general-purpose op-amp like the LM358 for a transresistance amplifier?
You can, but only for very low-speed, high-current applications where precision doesn't matter. The LM358 has a low gain-bandwidth product (~1 MHz) and a relatively high input bias current (~20 nA to 45 nA). If your sensor outputs 500 µA, the 45 nA bias current is a negligible 0.009% error. But if you are trying to measure a 100 nA signal from a UV photodiode, the LM358's bias current will consume nearly half your signal before the amplifier even sees it, and its bipolar input stage will inject significant current noise. For anything involving precision optical sensing, upgrade to a FET-input part like the TLV9042 (low speed/low power) or the OPA656 (high speed).
